[maemo-users] N900: Calibrating the accelerometers?

From: Marcel tanuva at googlemail.com
Date: Mon Mar 15 21:22:17 EET 2010
Am Montag, den 15.03.2010, 09:13 -0700 schrieb lfl at soeiro.com.br:
> >> When my accelerometer is working (it gets stuck every time the N900
> >> runs out of battery and is recharged) and I place it on a flat table,
> >> AccDisplay shows:
> >>
> >> X: 0.02
> >> Z: -0.95
> >> Y: 0.02
> >>
> >> I could find any position where all variables become close to 0.
> >> Anyway, I don't know yet how it works. I guess the variables might be
> >> related to the current angle on that axis, but the angle is expressed
> >> from -1.00 to 1.00.
> >>
> >> Since it is called an "accelerometer", isn't it supposed to tell the
> >> current acceleration in any axis?
> >
> > Yes, it is. Therefore it's correctly showing acceleration of -0.95m/s²
> > downwards - which is the accel/force that makes you stay on the ground.
> > I'd say the accelerometer is working perfectly well. (Implying the
> > 0.02m/s² on x/y axes are caused by jitter.)
> Hello Marcel
> Did you intend to reply only to me or to the maemo-list?

Sorry, I meant to reply to the list. CC'ed it now.

> Last time I checked, earth gravity's acceleration was 9.81m/s^2 ;-)

Sounds reasonable actually...

> Joking aside, if it is measuring acceleration indeed it must be in another
> unit, right?

If it's any unit at all. Maybe one could just convert it for human

> Those values of 0.02 can be attributed to either small vibrations in the
> building where the table is placed or just hardware fluctuation. I just
> would like to make sense of the readings.

If you get the same -0.95whateverunititis in every direction while
rotating the N900, I'd guess at least the hardware works okay.


More information about the maemo-users mailing list